lexoid 0.1.6.post1

No project description provided

Lexoid

Lexoid is an efficient document parsing library that supports both LLM-based and non-LLM-based (static) PDF document parsing.

Motivation:

• Use the multi-modal advancement of LLMs
• Enable convenience for users while driving innovation
• Collaborate with a permissive license

Installation

To install dependencies:

make install

or, to install with dev-dependencies:

make dev

To activate virtual environment:

source .venv/bin/activate

To use LLM-based parsing, define the following environment variables or create a .env file with the following definitions

OPENAI_API_KEY=""
GOOGLE_API_KEY=""

To build a .whl file for testing:

poetry build

Optionally, to use Playwright for retrieving web content with the .whl package (else regular requests will be used by default):

playwright install --with-deps --only-shell chromium

Usage

Example Notebook

Here's a quick example to parse documents using Lexoid:

from lexoid.api import parse
from lexoid.api import ParserType

parsed_md = parse("https://www.justice.gov/eoir/immigration-law-advisor", parser_type="LLM_PARSE", raw=True)
# or
pdf_path = "path/to/immigration-law-advisor.pdf"
parsed_md = parse(pdf_path, parser_type="LLM_PARSE", raw=True)

print(parsed_md)

Parameters

• path (str): The file path or URL.
• parser_type (str, optional): The type of parser to use ("LLM_PARSE" or "STATIC_PARSE"). Defaults to "AUTO".
• raw (bool, optional): Whether to return raw text or structured data. Defaults to False.
• pages_per_split (int, optional): Number of pages per split for chunking. Defaults to 4.
• max_threads (int, optional): Maximum number of threads for parallel processing. Defaults to 4.
• **kwargs: Additional arguments for the parser.

Benchmark

Initial results (more updates soon)

Rank	Model/Framework	Similarity	Time (s)
1	gpt-4o	0.799	21.77
2	gemini-1.5-pro	0.742	15.77
3	gpt-4o-mini	0.721	14.86
4	gemini-1.5-flash	0.702	4.56